|Marti J. Anderson (NZ)||email@example.com|
|Ecological data sets present special statistical problems. In particular, such data sets often consist of measurements or counts of individual species in an assemblage or community as a response to environmental or ecological factors. Problems arise in analysis that are specific to these kinds of data, but can be encountered in many other areas as well. To begin with, the number of species variables often exceeds the number of observations. In addition, the variables are discrete, rather than continuous, are truncated at zero (cannot take negative values) and often have highly skewed distributions. Such data sets also tend to have many zeros, due to species that are rare or patchily distributed. All of these characteristics of the data mean that traditional multivariate statistics cannot be applied in these systems. A further problem is that the data sets often arise in complex experimental designs. A challenge is to find robust and appropriate statistical approaches for such data sets. One possibility includes multivariate analyses based on distance functions. Here I will describe various new methods, based on distance matrices, which tackle these problematic statistical issues for modelling multivariate ecological data in complex experimental designs. I will also touch on the important aspects of teaching young statisticians or practicing ecologists what aspects of ecological data are unique and require special consideration for statistical analysis and modelling.|
|Download in Adobe Acrobat format (167 Kb).|
Back to top
ICOTS-6, The Sixth International Conference on Teaching Statistics - International Program Committee (IPC) Website.
Copyright © 2001 by the IASE. All rights reserved. This information is subject to change without notice. This page was last modified on July 2, 2002.
For questions or comments, contact the Webmaster, Dagan Ben-Zvi.